Determination of sugars in starch hydrolysates by IR spectroscopy

Sugar syrups are industrially obtained by enzymatic hydrolysis of starch and from aqueous mixtures of glucose, maltose (Dp2), maltotriose (Dp3), and other higher sugars (Dpn). In this work, new methods based on near-infrared (NIR) and mid-infrared (MIR) spectroscopy for analyzing starch hydrolysates were developed. Calibration models for the parameters of interest (viz. dextrose equivalent, dry matter, and individual sugars) were constructed using partial least-squares regression (PLSR). Various wavelength ranges and the absorbance, first-derivative and second-derivative spectral modes were tested in both spectral regions. The higher spectral correlation in the NIR region precluded the obtainment of acceptable PLSR-NIR models for all the constituents; in fact, the prediction errors for glucose, maltose, Dp3, and Dpn exceeded 7%, even though those for the dextrose equivalent and total sugars were only about 2%. The less extensive correlation between sugar signals in the MIR region allowed the development of models with much smaller errors for all the analytes.